Smaller Class Sizes Not the Answer

Graph credit: John Butcher

Parents and educators commonly believe that smaller class sizes provide a superior education. It seems logical: Smaller classes allow teachers to provide more individual attention to students. But the evidence supporting this intuitive view is surprisingly thin. A new study by the Campbell Collaboration, which promotes social and economic change through evidence-based policy, which calls the conventional wisdom into question.

The belief in the efficacy of small class sizes has inspired class-size reductions in the United States, Great Britain and Netherlands, observes the study, “Small class sizes for improving student achievement in elementary and secondary schools: a systematic review.” But the investment is expensive? Does it work?

The researchers conducted a meta-analysis of 127 studies that analyzed 55 student populations across 41 countries. The conclusion: “There is some evidence to suggest that there is an effect of reducing class size on reading achievement, though the effect is very small. There is no significant effect on mathematics achievement.”

While the overall effect is negligible, smaller class sizes have been shown to be more beneficial for students from socioeconomically disadvantaged backgrounds.

Smaller class sizes are costly, however, the study argues, and there may be more cost-effective strategies for improving student achievement.

I asked John Butcher, of Cranky’s Blog fame, to run an analysis for Virginia school systems. The chart above, which compares teacher-pupil ratios and math Standards of Learning pass rates of Virginia school districts, shows a negative (though statistically insignificant) correlation between teacher-pupil ratios and academic performance — totally consistent with the Campbell Collaboration study.

Bacon’s bottom line: Parents and educators in Virginia take it as an article of faith that class size is a key determinant of educational quality. But the effect may be far less significant than imagined. Instead of pushing for smaller classes, perhaps educators should consider different strategies.

For example: A judicious increase in class sizes — say, from 24 students per class to 26 students, should have no discernible effect on student achievement. But what if adjusting the teacher-student ratio freed resources that could be redirected to staffing classrooms dedicated to disadvantaged or disabled kids with a track record of disrupting their classes? Would such focused classes do a better job of helping the troubled students? Would the removal of classroom troublemakers allow teachers do a better job of teaching the other students?

I don’t know the answer. And even if I thought I did, I’d favor running pilot programs to test the idea before rolling it out on a large scale. But this is the kind of thinking, I submit, in which we need to engage rather than doubling down on stale and failed prescriptions.

There are currently no comments highlighted.

7 responses to “Smaller Class Sizes Not the Answer

  1. re: ”
    I don’t know the answer. And even if I thought I did, I’d favor running pilot programs to test the idea before rolling it out on a large scale. But this is the kind of thinking, I submit, in which we need to engage rather than doubling down on stale and failed prescriptions”

    was trying to find the academic background and qualifications of the authors like Trine Filges … who as far as I can tell is identified as a “Senior Researcher”. I think it’s important to know what researchers field of study is and in this case it seems to be:
    TITLE Senior Researcher
    DEPARTMENT SFI Campbell
    WORKS WITH Labour market, Impact measurement, Unemployment, Theory and methods

    The paper – to it’s credit does do a fairly thorough job of identifying the various factors that affect classroom size and learning.

    But the data slice presented here is basically a top level look which basically shows that the current class size is about equal to the current SOL results without ever really identifying factors and their involvement.

    There is no denying that some kinds of instruction to some kids needs to be with very small groups… especially for children who have learning issues.

    We also know that at the College level – freshmen often end up in very large classrooms with Graduate Assistant instructors and it’s sink or swim for them.

    But I just am troubled by taking the data provided by VDOE then concluding that it “shows that we might be able to bump up class size to free up money with precious little about how to do it other than it should be “pilot”.

    This study was essentially a literature review and other efforts here seem to focus primarily on public education but I would suggest that we actually have ways to compare and two are:

    1. – non-public schools – how does class size work for them

    2. – computer/online instruction which is rapidly advancing and is getting to the point where it can and does function for some kids as one-on-one tutoring.

    In fact, I would expect the non-public schools who are held to a higher productivity standard for money verses effort might well be making inroads worth knowing about.

    I fundamentally believe in the concept of public education but I also acknowledge that it has a lot of warts and guaranteed tenure and funding are impediments to change and to motivation to change and that’s especially true in low-income neighborhood schools that typically are not well resourced in terms of experienced staff. Our current system of neighborhood schools is an unmitigated disaster when it comes to disadvantaged kids living in low income neighborhoods is a failure and public education needs competitors in this area.

    Having said that – I just don’t think the current process of taking the data that public schools provide as part of their transparency efforts and using it to impugn them on their “failures” is helpful nor productive especially when suggestions for “reform” are more often than not just vague as hell.

    We spend too much time tearing down our institutions these days and precious little time on task to reform and improve… an optimistic view of change for the better if you will.

  2. On this one I consulted the Middle School Math Teacher of the Year, since that expert shares my life. Class size is not irrelevant, and 10 is obviously better for the students than 40, but the research is clear – has been for a long time, she said – that class size is not the key variable or even a major one. Except of course for the really key class size of one student and one parent huddled over formal homework or some other activity where math is a key element. MSMTY mentioned above can turn a trip to Kroger into a math lesson, and I can remember my father making me compute distances, speeds and time in my head using a map on long car trips. No GPS in 1967! He forced me to estimate – 30 minutes to that next town.

    Interesting how you brought this discussion back to your favorite hobby horse of disruptive students. Pick up the phone and set up a visit to any school district and find out what they are doing – you might be surprised.

  3. Pingback: Larger Classes, Same Performance – CrankysBlog

  4. Oops!

    I used the 2016 pass rates and the 2017 teacher/student ratios in that graph. I’ve posted the correct graph (and its siblings for the other four subjects) here:
    https://calaf.org/?p=6406

  5. My hero Thomas Sowell endorses the judgement of the Middle School Math Teacher of the Year who opines that “Class size is not irrelevant, and 10 is obviously better for the students than 40, but the research is clear – has been for a long time, she said – that class size is not the key variable or even a major one. Except of course for the really key class size of one student and one parent huddled over formal homework or some other activity where math is a key element …”

    In addition, I strongly suspect that the Middle School Math Teacher of the Year knows that a single teacher can step out alone with a single student, from time to time, and change that struggling student’s academic life, as teacher Laurence ‘Larry” Turner did for me long ago in the 11th grade.

Leave a Reply